Optimal network size for most robust Turing patterns

This repository contains notebooks and Python scripts used to generate figures 2-8 in the following paper.

The code was developed by Dr. Robert Endres' group at Imperial College London.

---

Code organisation:

• figure_5.jpynb - tutorial on how to generate circular scatter plot and corresponding kernel density estimations of eigenvalues in the complex plan with examples
• figure_6.jpynb - tutorial on how to generate:
  • a) % of stable random matrices vs radius γ plot
  • b) % of stable random matrices turning unstable with diffusion vs radius γ plot
  • c) % of Turing instabilities vs raidus γ plot
  • d) % of Turing 1 instabilities vs raidus γ plot
• figure_7.jpynb - comprehensive explanation of figure7_re_v5.py
• figure7_re_v5.py - generates heatmaps of % Turing 1 occurences and corresponding percentage shares plots
• figure_8.jpynb - comprehensive explanation of figure8_RE_v3.py
• figure8_RE_v3.py - generates heatmaps of % Turing 1 occurences for different diffusion parameters D and network sizes N

---

Requirements

In order to run the Python scripts you need:

---

Contacts

Dr. Robert Endres: r.endres@imperial.ac.uk